Deformation recognition method, device and electronic equipment

Abstract

The present invention provides a deformation recognition method, device and electronic device, which relate to the field of image processing technology. The method includes: obtaining at least two groups of three-dimensional point cloud data of a target object; wherein different groups of three-dimensional point cloud data correspond to different states of the target object; performing compression processing on each group of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data; performing conversion calculation processing on the compressed three-dimensional point cloud data to obtain two-dimensional grid data; performing division processing on the compressed three-dimensional point cloud data to obtain a plurality of divided first data units containing the three-dimensional point cloud data; performing deformation recognition processing based on the relevant data of the first data unit and the two-dimensional grid data to determine the deformation result. The solution of the present invention solves the problem that the current deformation recognition method cannot meet the demand of quickly and accurately identifying small deformations.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a deformation recognition method, device and electronic equipment. Background Art

[0002] Spatial geometric deformation recognition has a wide range of applications, such as facial micro-expression recognition and material processing (such as vehicle shell molding and mechanical modules). However, current spatial geometric deformation recognition methods are generally based on two-dimensional image processing, which cannot meet the requirements of quickly and accurately identifying small deformations. Summary of the Invention

[0003] The purpose of the present invention is to provide a deformation recognition method, device and electronic equipment to solve the problem that the current deformation recognition method cannot meet the demand of quickly and accurately identifying small deformations.

[0004] To achieve the above objectives, an embodiment of the present invention provides a deformation recognition method, comprising:

[0005] Acquire at least two sets of three-dimensional point cloud data of a target object; wherein different sets of three-dimensional point cloud data correspond to different states of the target object;

[0006] Performing compression processing on each set of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data;

[0007] Perform conversion and calculation on the compressed three-dimensional point cloud data to obtain two-dimensional grid data;

[0008] performing segmentation processing on the compressed three-dimensional point cloud data to obtain a plurality of segmented first data units containing the three-dimensional point cloud data;

[0009] Deformation recognition processing is performed based on the relevant data of the first data unit and the two-dimensional grid data to determine a deformation result.

[0010] Optionally, obtaining at least two sets of three-dimensional point cloud data of the target object includes:

[0011] For the target object in different states, the target object is scanned multiple times by a 3D scanning device and the point cloud data obtained by the scanning is preprocessed to obtain a set of three-dimensional point cloud data corresponding to the target object in each state.

[0012] Optionally, the compressing each set of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data includes:

[0013] Dividing the target three-dimensional point cloud data to obtain a plurality of second data units containing three-dimensional point cloud data; wherein the target three-dimensional point cloud data is any one set of three-dimensional point cloud data among the at least two sets of three-dimensional point cloud data;

[0014] calculating a curvature corresponding to the second data unit;

[0015] The three-dimensional point cloud data in the second data unit is compressed according to the curvature corresponding to the second data unit to obtain compressed three-dimensional point cloud data.

[0016] Optionally, calculating the curvature corresponding to the second data unit includes:

[0017] Determining a target plane based on any three points contained in a target data unit; wherein the target data unit is any one of the plurality of second data units;

[0018] Calculating the distance between each point contained in the target data unit and the target plane;

[0019] A standard deviation corresponding to the distance between each point and the target plane is calculated, and the standard deviation is determined as the curvature.

[0020] Optionally, compressing the three-dimensional point cloud data in the second data unit according to the curvature corresponding to the second data unit to obtain compressed three-dimensional point cloud data includes:

[0021] For a second data unit whose curvature is smaller than a first threshold, retaining a point in the second data unit that is closest to the center of the second data unit;

[0022] For a second data unit having a curvature greater than or equal to a first threshold, retaining a plurality of points in the second data unit based on a curvature reduction algorithm;

[0023] The points retained in each second data unit are determined as compressed three-dimensional point cloud data.

[0024] Optionally, the second data unit is a cuboid containing three-dimensional point cloud data, and the first threshold is determined based on the minimum value of the length, width and height of the cuboid.

[0025] Optionally, performing conversion calculation processing on the compressed three-dimensional point cloud data to obtain two-dimensional grid data includes:

[0026] Performing a first projection process on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data;

[0027] Based on a local greedy projection triangulation algorithm, triangular mesh data with a connection relationship is obtained from the two-dimensional point cloud data;

[0028] Based on the triangular mesh data, the number of triangular meshes and the total area of ​​the triangular meshes are determined, and the two-dimensional mesh data is determined according to the number of triangular meshes and the total area of ​​the triangular meshes.

[0029] Optionally, performing a first projection process on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data includes:

[0030] Determining the center coordinates of the compressed three-dimensional point cloud data in a three-dimensional coordinate system;

[0031] Based on the center coordinates, dividing the compressed three-dimensional point cloud data into a plurality of third data units containing three-dimensional point cloud data;

[0032] determining a projection surface based on each of the third data units, and establishing a two-dimensional coordinate system based on the projection surface;

[0033] Determine the three-dimensional projection coordinates of each point in the third data unit on the projection plane corresponding to the third data unit;

[0034] Based on the two-dimensional coordinate system and the three-dimensional projection coordinates, the two-dimensional coordinates corresponding to the points in each third data unit are determined, and the projected two-dimensional point cloud data is determined based on the two-dimensional coordinates.

[0035] Optionally, performing deformation recognition processing based on the relevant data of the first data unit and the two-dimensional grid data to determine the deformation result includes:

[0036] determining a degree of deformation according to the number of triangular meshes in the two-dimensional mesh data corresponding to the three-dimensional point cloud data;

[0037] determining a deformation rate according to a total area of ​​a triangular mesh in the two-dimensional network data corresponding to the three-dimensional point cloud data;

[0038] determining a deformation position according to the number of points included in the first data unit corresponding to the three-dimensional point cloud data;

[0039] A deformation result is determined according to the deformation degree, the deformation rate, and the deformation position.

[0040] To achieve the above objectives, an embodiment of the present invention provides a deformation recognition device, comprising:

[0041] An acquisition module, configured to acquire at least two sets of three-dimensional point cloud data of a target object; wherein different sets of three-dimensional point cloud data correspond to different states of the target object;

[0042] The first processing module is used to compress each set of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data;

[0043] The second processing module is used to convert and calculate the compressed three-dimensional point cloud data to obtain two-dimensional grid data;

[0044] a third processing module, configured to divide the compressed three-dimensional point cloud data to obtain a plurality of divided first data units containing the three-dimensional point cloud data;

[0045] The fourth processing module is configured to perform deformation recognition processing based on the relevant data of the first data unit and the two-dimensional grid data to determine a deformation result.

[0046] Optionally, the acquisition module includes:

[0047] The acquisition submodule is used to scan the target object in different states multiple times using a 3D scanning device and pre-process the point cloud data obtained by the scans to obtain a set of three-dimensional point cloud data corresponding to the target object in each state.

[0048] Optionally, the first processing module includes:

[0049] a first segmentation submodule, configured to segment the target three-dimensional point cloud data to obtain a plurality of second data units containing three-dimensional point cloud data; wherein the target three-dimensional point cloud data is any one set of three-dimensional point cloud data among the at least two sets of three-dimensional point cloud data;

[0050] A first calculation submodule, configured to calculate the curvature corresponding to the second data unit;

[0051] The compression submodule is used to compress the three-dimensional point cloud data in the second data unit according to the curvature corresponding to the second data unit to obtain compressed three-dimensional point cloud data.

[0052] Optionally, the first calculation submodule includes:

[0053] a first determining unit, configured to determine a target plane based on any three points contained in a target data unit, wherein the target data unit is any one of the plurality of second data units;

[0054] a first calculation unit, configured to calculate the distance between each point included in the target data unit and the target plane;

[0055] The second calculation unit is configured to calculate a standard deviation corresponding to the distance between each point and the target plane, and determine the standard deviation as the curvature.

[0056] Optionally, the compression submodule includes:

[0057] a first compression unit, configured to retain, for a second data unit having a curvature smaller than a first threshold, a point in the second data unit that is closest to a center of the second data unit;

[0058] a second compression unit, configured to retain, for a second data unit having a curvature greater than or equal to a first threshold, a plurality of points in the second data unit based on a curvature reduction algorithm;

[0059] The second determining unit is configured to determine the points retained in each second data unit as compressed three-dimensional point cloud data.

[0060] Optionally, the second data unit is a cuboid containing three-dimensional point cloud data, and the first threshold is determined based on the minimum value of the length, width and height of the cuboid.

[0061] Optionally, the second processing module includes:

[0062] A first projection submodule is used to perform a first projection process on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data;

[0063] A second projection submodule is configured to obtain triangular mesh data having a connection relationship from the two-dimensional point cloud data based on a local greedy projection triangulation algorithm;

[0064] The determination submodule is configured to determine the number of triangular meshes and the total area of ​​the triangular meshes based on the triangular mesh data, and determine the number of triangular meshes and the total area of ​​the triangular meshes as the two-dimensional mesh data.

[0065] Optionally, the first projection submodule includes:

[0066] a third determining unit, configured to determine the center coordinates of the compressed three-dimensional point cloud data in a three-dimensional coordinate system;

[0067] a dividing unit, configured to divide the compressed three-dimensional point cloud data into a plurality of third data units containing three-dimensional point cloud data based on the center coordinates;

[0068] an establishing unit, configured to determine a projection surface based on each of the third data units, and establish a two-dimensional coordinate system based on the projection surface;

[0069] a fourth determining unit, configured to determine the three-dimensional projection coordinates of each point in the third data unit on the projection plane corresponding to the third data unit;

[0070] The fifth determining unit is used to determine the two-dimensional coordinates corresponding to the points in each third data unit based on the two-dimensional coordinate system and the three-dimensional projection coordinates, and determine the projected two-dimensional point cloud data based on the two-dimensional coordinates.

[0071] Optionally, the fourth processing module includes:

[0072] A first processing submodule is configured to determine a degree of deformation according to the number of triangular meshes in the two-dimensional mesh data corresponding to the three-dimensional point cloud data;

[0073] a second processing subunit, configured to determine a deformation rate according to a total area of ​​a triangular mesh in the two-dimensional network data corresponding to the three-dimensional point cloud data;

[0074] a third processing submodule, configured to determine a deformation position according to the number of points contained in the first data unit corresponding to the three-dimensional point cloud data;

[0075] The fourth processing submodule is configured to determine a deformation result according to the deformation degree, the deformation rate, and the deformation position.

[0076] To achieve the above-mentioned purpose, an embodiment of the present invention provides an electronic device, including a transceiver, a processor, a memory, and a program or instruction stored in the memory and executable on the processor; when the processor executes the program or instruction, the steps in the deformation recognition method described above are implemented.

[0077] To achieve the above objectives, an embodiment of the present invention provides a readable storage medium having a program or instruction stored thereon, which implements the steps of the above-mentioned deformation recognition method when executed by a processor.

[0078] The beneficial effects of the above technical solution of the present invention are as follows:

[0079] The embodiments of the present invention can reduce the computational complexity of subsequent processing and improve processing efficiency by acquiring three-dimensional point cloud data of a target object and compressing the acquired three-dimensional point cloud data. Furthermore, the compressed three-dimensional point cloud data is converted and computed to obtain two-dimensional grid data. The compressed three-dimensional point cloud data is divided and processed to obtain a plurality of first data units containing the three-dimensional point cloud data. Deformation recognition is performed based on the relevant data of the first data units and the two-dimensional grid data to determine the deformation result. This solves the current problem of being unable to perform deformation recognition on three-dimensional images. Compared with current deformation recognition methods based on two-dimensional images, this method can meet the needs of quickly and accurately identifying small deformations. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 is a flow chart of a deformation recognition method according to an embodiment of the present invention;

[0081] Figure 2 A schematic diagram of a network topology for deformation recognition according to an embodiment of the present invention;

[0082] Figure 3 is a block diagram of a deformation recognition device according to an embodiment of the present invention;

[0083] Figure 4 FIG. 4 is a block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0084] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0085] It should be understood that references throughout this specification to "one embodiment" or "an embodiment" mean that a particular feature, structure, or characteristic associated with the embodiment is included in at least one embodiment of the present invention. Therefore, the appearances of "in one embodiment" or "in an embodiment" throughout this specification do not necessarily refer to the same embodiment. Furthermore, these particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0086] In various embodiments of the present invention, it should be understood that the size of the serial numbers of the following processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0087] Additionally, the terms "system" and "network" are often used interchangeably herein.

[0088] In the embodiments provided herein, it should be understood that "B corresponding to A" means that B is associated with A and B can be determined based on A. However, it should also be understood that determining B based on A does not mean determining B based solely on A; B can also be determined based on A and / or other information.

[0089] like Figure 1 As shown, an embodiment of the present invention provides a deformation recognition method, including:

[0090] Step 11: Obtain at least two sets of three-dimensional point cloud data of the target object; wherein different sets of three-dimensional point cloud data correspond to different states of the target object.

[0091] The target object can be understood as any object whose deformation is to be identified.

[0092] Optionally, a set of three-dimensional point cloud data may correspond to a state of a target object. The different states of the target object may refer to different moments of the target object, or may refer to deformation states of the target object (such as multiple states before and after deformation, or during deformation).

[0093] Step 12: Perform compression processing on each set of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data.

[0094] Step 13: Perform conversion and calculation on the compressed three-dimensional point cloud data to obtain two-dimensional grid data.

[0095] Step 14: Divide the compressed three-dimensional point cloud data to obtain a plurality of divided first data units containing the three-dimensional point cloud data.

[0096] Step 15: Perform deformation recognition processing based on the relevant data of the first data unit and the two-dimensional grid data to determine a deformation result.

[0097] In the above scheme, by acquiring the three-dimensional point cloud data of the target object and compressing the acquired three-dimensional point cloud data, the computational complexity of subsequent processing can be reduced and processing efficiency can be improved. Furthermore, the compressed three-dimensional point cloud data is converted and calculated to obtain two-dimensional grid data; the compressed three-dimensional point cloud data is divided and processed to obtain multiple first data units containing the three-dimensional point cloud data. Deformation recognition is performed based on the relevant data of the first data units and the two-dimensional grid data to determine the deformation result. This solves the current problem of being unable to perform deformation recognition on three-dimensional images. Compared with current deformation recognition methods based on two-dimensional images, it can meet the needs of quickly and accurately identifying small deformations.

[0098] like Figure 2 As shown, an embodiment of the present invention also provides a network topology diagram for deformation recognition, which includes: a 3D scanning module, a preprocessing module, a data compression module, a network generation module, a calculation module, and an analysis and processing module. The following describes the deformation recognition method of the embodiment of the present invention in detail, combining the above network topology.

[0099] Optionally, obtaining at least two sets of three-dimensional point cloud data of the target object includes:

[0100] For the target object in different states, the target object is scanned multiple times by a 3D scanning device and the point cloud data obtained by the scanning is preprocessed to obtain a set of three-dimensional point cloud data corresponding to the target object in each state.

[0101] In this embodiment, a 3D scanning device with 3D scanning and pre-processing functions can be used to obtain 3D point cloud data of the target object in different states. For example, the 3D scanning device can implement the functions of the above-mentioned 3D scanning module and pre-processing module.

[0102] The 3D scanning module is a high-precision object measurement device. It continuously projects a grating onto the surface of an object, while the camera simultaneously captures images. The images are then calculated and the three-dimensional spatial coordinates (X, Y, Z) of the two images are calculated using phase-stability polar lines. This allows the measurement of the three-dimensional contour of the object's surface. The point cloud data obtained through 3D scanning can be a set of points including their three-dimensional coordinates and color attributes.

[0103] Since it is difficult for a 3D scanning device (such as a 3D scanner) to obtain complete point cloud data of a scanned target (i.e., a target object) by scanning once from one direction, it means that the entity information of the scanned target object usually needs to be completed through several scans. However, each scanned image frame is a local coordinate system with the scanner position as the zero point, that is, the coordinate system of the point cloud data obtained from each scan is independent and unrelated. In addition, since each point cloud array data is actually part of the scanned scene, it is necessary to convert these point cloud array data into the same coordinate system. Therefore, the obtained point cloud data should be spliced ​​and matched. In addition, due to the complex acquisition environment, different image brightness and darkness, and the influence of the uneven reflection characteristics of the scanned target object itself, data unrelated to the graphics will be obtained, and there will be certain noise and interference. Therefore, the point cloud data obtained by the 3D scanning device needs to be preprocessed. This process can be implemented by a preprocessing module.

[0104] Among them, the preprocessing module: usually reprocesses the collected raw data, checks the integrity and consistency of the data, standardizes the data format, and filters the point cloud. Denoising, deletion and reduction of point cloud density will also be performed during the processing. This process is also called point cloud filtering, which is a very important process in preprocessing. The preprocessing module that comes with the 3D scanner can be used here. In this embodiment, the data processed by the 3D scanner can be used directly. The 3D scanner data contains point cloud coordinates, color, reflection intensity and other information. The initialization module here only filters the coordinate position information related to the present invention, such as P(P_i∈P)(x, y, z) coordinates. The filtering here can directly use the text tool to extract the corresponding columns. Optionally, the output data of the preprocessing module can be a data table table_init(x, y, z) of the initial point coordinates.

[0105] Optionally, the three-dimensional point cloud data obtained by scanning is compressed by a data compression module to obtain compressed three-dimensional point cloud data. The data compression module is used to compress each set of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data, which may include:

[0106] Dividing the target three-dimensional point cloud data to obtain a plurality of second data units containing three-dimensional point cloud data; wherein the target three-dimensional point cloud data is any one set of three-dimensional point cloud data among the at least two sets of three-dimensional point cloud data;

[0107] Calculating the curvature corresponding to the second data unit (eg, calculating the curvature corresponding to each second data unit respectively);

[0108] According to the curvature corresponding to the second data unit, the three-dimensional point cloud data in the second data unit is compressed to obtain compressed three-dimensional point cloud data (e.g., according to the curvature corresponding to each second data unit, the three-dimensional point cloud data in the second data unit is compressed to obtain compressed three-dimensional point cloud data).

[0109] In this embodiment, the point cloud data obtained during the acquisition process of the 3D point cloud data is very large, and the subsequent processing will be very computationally intensive and inefficient. Therefore, the method of compressing the 3D point cloud data in this solution can not only improve the computational efficiency but also well preserve the geometric features of the object.

[0110] Optionally, the second data unit may be a cuboid containing three-dimensional point cloud data.

[0111] For example, the three-dimensional point cloud data obtained after preprocessing can be divided into rectangular regions. Here, the length, width, and height of the rectangular region obtained by division can be represented by l, m, and n. Optionally, the length, width, and height of the rectangular region can be pre-set or determined based on model training and learning, but the embodiments of the present application are not limited to this.

[0112] Optionally, calculating the curvature corresponding to the second data unit includes:

[0113] Determining a target plane based on any three points contained in a target data unit; wherein the target data unit is any one of the plurality of second data units;

[0114] Calculating the distance between each point contained in the target data unit and the target plane;

[0115] A standard deviation corresponding to the distance between each point and the target plane is calculated, and the standard deviation is determined as the curvature.

[0116] For example: First, sort the obtained cuboids (i.e., the second data units). For each cuboid, select any three points contained therein to calculate the plane equation (i.e., the plane equation corresponding to the target plane), and then calculate the standard deviation σ of the average distance from all points to this plane equation. This standard deviation σ can be the key for subsequent curvature judgment. For example, the curvature can be approximated based on the calculated standard deviation.

[0117] Among them, the standard deviation can be calculated by the plane fitting method, and the calculation formula of this standard deviation is as follows:

[0118]

[0119] Among them, N is the total number of points contained in the cuboid; μ is the average value of the distances from all points in the cuboid to the target plane; x i is the distance from the i-th point in the cuboid to the target plane, where i is a positive integer and i is less than or equal to N.

[0120] Optionally, the compression processing of the three-dimensional point cloud data in the second data unit according to the curvature corresponding to the second data unit to obtain the compressed three-dimensional point cloud data includes:

[0121] For the second data unit with a curvature less than the first threshold, retain one point closest to the center of the second data unit in the second data unit;

[0122] For the second data unit with a curvature greater than or equal to the first threshold, retain multiple points in the second data unit based on the curvature reduction algorithm;

[0123] Determine the points retained in each second data unit as the compressed three-dimensional point cloud data.

[0124] Optionally, the first threshold can be set in advance according to the system. For example, the first threshold is determined based on the minimum value of the length, width, and height of the cuboid. This first threshold can also be called the judgment value S. The half of the minimum value of the length, width, and height of the cuboid can be determined as the judgment value S, that is, S = 1 / 2min(l, m, n). Of course, the judgment value S can also be comprehensively considered according to the object smoothness, calculation time, and feature retention degree. The embodiments of the present invention are not limited thereto.

[0125] Further, for the cuboid with a relatively small curvature, that is, the second data unit with a curvature less than the first threshold (σ < S), the bounding box method can be used, and each cuboid retains one point closest to the center; for the cuboid with a relatively large curvature, that is, the second data unit with a curvature greater than or equal to the first threshold (σ ≥ S), the curvature reduction method can be used, and each cuboid retains multiple points.

[0126] Among them, the curvature simplification algorithm can be to use the least squares method to fit the surface and then determine whether to delete the point according to the curvature simplification principle. It means that the point is relatively flat and should be deleted; i For any point in the cuboid, is the geometric center of the cuboid, such as The coordinates are in is the average value of the X-axis coordinates of all points in the cuboid, is the average value of the Y-axis coordinates of all points in the cuboid, is the average value of the Z-axis coordinates of all points in the cuboid.

[0127] The fitting equation is:

[0128]

[0129] Among them, c i,j-i is the coefficient of the term in the fitting equation.

[0130] In this embodiment, the cuboids obtained by division are pre-judged during simplification, and the judgment values ​​can be pre-set based on the specific shape through machine learning. This is relatively flexible and adaptable to a variety of objects. A corresponding simplification algorithm is then employed, which maximizes the preservation of the object's features compared to a simple bounding box method and significantly improves the calculation speed compared to a simple curvature simplification method. Optionally, the input data for the data compression module is table_init(x,y,z),m,n,l,S; the output data of the data compression module is a table of compressed point coordinates, table_compress(id,x,y,z), where id is the number of the point in the point cloud data.

[0131] Optionally, the step of converting and calculating the compressed three-dimensional point cloud data to obtain two-dimensional grid data may be performed by a grid generation module and a calculation module to convert and calculate the compressed three-dimensional point cloud data to obtain two-dimensional grid data; and the step of dividing the compressed three-dimensional point cloud data to obtain a plurality of divided first data units containing three-dimensional point cloud data may be performed by a grid generation module and an analysis and processing module to divide the compressed three-dimensional point cloud data to obtain a plurality of divided first data units containing three-dimensional point cloud data.

[0132] Optionally, performing conversion calculation processing on the compressed three-dimensional point cloud data to obtain two-dimensional grid data includes:

[0133] Performing a first projection process on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data;

[0134] Based on a local greedy projection triangulation algorithm, triangular mesh data with a connection relationship is obtained from the two-dimensional point cloud data;

[0135] Based on the triangular mesh data, the number of triangular meshes and the total area of ​​the triangular meshes are determined, and the two-dimensional mesh data is determined according to the number of triangular meshes and the total area of ​​the triangular meshes.

[0136] For example, a mesh generation module performs a first projection on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data; and a local greedy projection triangulation algorithm is used to obtain triangular mesh data with a connected relationship from the two-dimensional point cloud data. A calculation module calculates the number of triangular meshes and the total area of ​​the triangular meshes based on the triangular mesh data.

[0137] When converting point cloud data through the mesh generation module, the point cloud data can be numbered sequentially. Then, the numbered point cloud data is triangulated using the local greedy projection method to obtain the point cloud numbers of all triangles, that is, the triangular mesh data with a connected relationship. The input data of this mesh generation model is table_compress(id,x,y,z).

[0138] Optionally, performing a first projection process on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data includes:

[0139] Determining the center coordinates of the compressed three-dimensional point cloud data in a three-dimensional coordinate system;

[0140] Based on the center coordinates, dividing the compressed three-dimensional point cloud data into a plurality of third data units containing three-dimensional point cloud data;

[0141] determining a projection surface based on each of the third data units, and establishing a two-dimensional coordinate system based on the projection surface;

[0142] Determine the three-dimensional projection coordinates of each point in the third data unit on the projection plane corresponding to the third data unit;

[0143] Based on the two-dimensional coordinate system and the three-dimensional projection coordinates, the two-dimensional coordinates corresponding to the points in each third data unit are determined, and the projected two-dimensional point cloud data is determined based on the two-dimensional coordinates.

[0144] Specifically, we can first divide the area and project it: For example, we can calculate the center coordinates of the compressed 3D point cloud data in the 3D coordinate system using the following formula:

[0145]

[0146]

[0147]

[0148] Among them, x i is the X-axis coordinate of the i-th point in the compressed 3D point cloud data, y i is the Y-axis coordinate of the i-th point in the compressed 3D point cloud data, z i is the Z-axis coordinate of the i-th point in the compressed 3D point cloud data; n is the total number of points in the compressed 3D point cloud data, i is a positive integer, and i is less than or equal to n.

[0149] Based on the determined center coordinates Divide the input data table_compress(id,x,y,z) of the grid generation model into 8 tables, with the division range as follows:

[0150] table_son1:

[0151] table_son2:

[0152] table_son3:

[0153] table_son4:

[0154] table_son5:

[0155] table_son6:

[0156] table_son7:

[0157] table_son8:

[0158] Projection surfaces are made along 8 directions (i.e. the 8 areas obtained by the above division) Among them, the 8 projection surfaces are:

[0159]

[0160]

[0161]

[0162]

[0163]

[0164]

[0165]

[0166]

[0167] The plane equation corresponding to the projection surface is:

[0168]

[0169] Right now,

[0170] make, That is, we get the standard plane equation: ax+by+cz+d=0.

[0171] According to the projection formula from point to plane, the coordinate P of the i-th point (which can also be understood as any point) in the compressed 3D point cloud data is i (x i ,y i ,z i ), the corresponding projection point coordinates are: P i ′(x′ i ,y′ i ,z′ i ).

[0172] in,

[0173]

[0174] x′ i =z i -a*t

[0175] y′ i =z i -a*t

[0176] z′ i =z i -a*t

[0177] Then convert the projection coordinates into two-dimensional coordinates: For example, arrange the projection coordinates in reverse order of the X-axis coordinates, with the first point as the distance to the center point as the X-axis, and the perpendicular point as the Y-axis to establish a rectangular coordinate system, and convert the three-dimensional coordinates into two-dimensional coordinates. The specific method is as follows:

[0178] Select the coordinates of the first point as P′0(x′0,y′0,z′0) and establish a rectangular coordinate system on the plane;

[0179] Among them, the X-axis vector is:

[0180] The Y-axis vector is:

[0181] Traverse all points in table_compress(id,x,y,z) and reduce the three-dimensional coordinates to two-dimensional coordinates (x″) i , y″ i ):

[0182] P i ′(x′ i ,y′ i ,z′ i )→P i ″(x″ i ,y″ i )

[0183]

[0184]

[0185] In this embodiment, the input data table_compress(id,x,y,z) is processed to reduce the three-dimensional coordinates into two-dimensional coordinates to obtain an input two-dimensional list array: table_vex(id,x,y), that is, the projected two-dimensional point cloud data is determined based on the two-dimensional coordinates.

[0186] Optionally, based on a local greedy projection triangulation algorithm, obtaining triangular mesh data with a connection relationship through the two-dimensional point cloud data may include: generating a network process and processing a mesh boundary.

[0187] The network generation process includes, for example, using a local greedy projection triangulation method to generate a mesh, performing greedy projection triangulation on the data of the eight regions, such as using a conventional triangulation algorithm (Delaunay).

[0188] Mesh boundary processing: The mesh boundaries require some preprocessing. You can select coordinate point cloud data near the X, Y, and Z axes. For example, select the region (x-5l, x+5l), (y-5m, y+5m), and (z-5n, z+5n). Continue using the greedy projection algorithm for this region, replacing the previous mesh points with data from the region (x-3l, x+3l), (y-3m, y+3m), and (z-3n, z+3n).

[0189] In this embodiment, the greedy projection algorithm used by the mesh generation module is for closed figures in space and is in a real-time changing process. It can obtain the projection figure within the maximum range through multiple projections, including the processing of boundaries. Compared with the traditional greedy projection algorithm for objects with only one projection surface and no intersection, it ensures both calculation speed and mesh connection accuracy. Optionally, the data output of the mesh generation module is a triangular mesh with a connection relationship, and two tables are output: one table shows the triangular mesh with a connection relationship, table_mesh(id_a,id_b,id_c), where id_a, id_b, id_c are the numbers of the points that constitute the triangular mesh; the other table is table_coord(id,x,y,z), which can also be understood as the three-dimensional point cloud data after the above compression processing.

[0190] After the mesh generation module outputs a connected triangular mesh, the calculation module calculates the number and area of ​​the generated triangular meshes based on the point cloud number data output by the network generation module. The sum of the areas of all triangles can be used to approximate the surface area of ​​the target object. Optionally, the calculation module's input data is table_mesh(id_a, id_b, id_c) and table_coord(id, x, y, z), and the output data is the surface area S.

[0191] Alternatively, the surface area can be calculated using Heron's formula:

[0192] A(x1,y1,z1),B(x2,y2,z2),C(x3,y3,z3)

[0193]

[0194]

[0195]

[0196] p=(a+b+c) / 2

[0197]

[0198] Wherein, A, B, and C are three points constituting a triangular mesh; and S is the surface area, i.e., the total area of ​​the triangular mesh.

[0199] After the mesh generation module outputs a triangular mesh with a connection relationship, the analysis and processing module divides the area along the z-axis based on the point cloud data and compares it with the initial model. The purpose of the division is to determine the specific location of the deformation based on the change in the number of points contained in the point cloud data in the area.

[0200] Optionally, the division rules can be adjusted according to the actual target, and the size of the divided area can also be comprehensively considered based on the precision requirements and time efficiency, but the embodiments of the present invention are not limited to this. Once the rules and parameters are selected, they will no longer be changed subsequently to ensure that the three-dimensional point cloud data obtained for the target object in different states are consistent during processing, thereby ensuring the accuracy of the deformation recognition results. The division method here is similar to the cloud point compression algorithm, but the division method here is fixed after the initial division is completed (such as it can be understood as the processing of the three-dimensional point cloud data corresponding to the first state).

[0201] Taking the initial state as an example, the division to obtain a plurality of first data units containing three-dimensional point cloud data in this embodiment is described as follows:

[0202] According to the data of table_coord(id,x,y,z), we get P1(P1 i ∈P1), that is, the compressed three-dimensional point cloud data corresponding to the initial state, searches for the maximum and minimum values ​​of all points in P1 in the three-dimensional direction, which are: min ,x max ,y min ,y max ,z min ,z max Determine the smallest rectangular parallelepiped containing the point (here for the initial state, it can be understood as dividing according to the smallest rectangle containing the point, and for the state after the initial state, it is divided in the same way as the initial state, regardless of whether it contains the point) and divide it along the X-axis direction, Y-axis direction, and Z-axis direction to obtain (m1×n1×l1) small basic rectangular parallelepiped units (i.e., the first data unit). The sizes of the three sides of the basic rectangular parallelepiped unit are g1 and g2 respectively. x ,g1 y ,g1 z The output here is id1, x1, y1, z1, number, which is stored as table_cuboid_cell(); where id1 is the number of the basic cuboid cell after sorting by Z-axis coordinate, x1, y1, z1 are the coordinates of the center point of the basic cuboid cell respectively, and number is the number of points contained in the basic cuboid cell. Optionally, the input data of the analysis and processing module is table_coord(id, x, y, z), m1, n1, l1, and the output data of the analysis and processing module is table_cuboid_cell(id1, x1, y1, z1, number).

[0203] Optionally, after determining the number of triangular meshes and the total area of ​​the triangular meshes through the calculation module, and determining the number of points contained in each first data unit (ie, each basic rectangular unit cell) through the analysis and processing module, the above data is stored through the data storage module.

[0204] Optionally, performing deformation recognition processing based on the relevant data of the first data unit and the two-dimensional grid data to determine a deformation result includes:

[0205] determining a degree of deformation according to the number of triangular meshes in the two-dimensional mesh data corresponding to the three-dimensional point cloud data;

[0206] determining a deformation rate according to a total area of ​​a triangular mesh in the two-dimensional network data corresponding to the three-dimensional point cloud data;

[0207] determining a deformation position according to the number of points included in the first data unit corresponding to the three-dimensional point cloud data;

[0208] A deformation result is determined according to the deformation degree, the deformation rate, and the deformation position.

[0209] Specifically, for each state of the target object (e.g., each time a shape deforms, or at different times), a corresponding number of triangles and surface area are generated. The magnitude and rate of deformation can be measured by the magnitude of these fluctuations. The specific location of the deformation can also be determined by comparing the number of points contained in each basic rectangular unit cell. The deformation rate is the ratio of the surface area after deformation to the surface area of ​​the initial model.

[0210] Since the position of the 3D scanning device is fixed during the deformation test, that is, the reference point of the coordinate system remains unchanged, and the only thing that changes is the deformed part of the object. Therefore, the deformation position and size of the object can be determined according to the above method. The object deformation recognition process can be expressed as:

[0211] Deformation state i: / / i represents the i-th state of the object deformation collection;

[0212] Input: table_coord(id,x,y,z)

[0213] Output:table_deform(id,number)

[0214] If x∈(x1-g1 x / 2,x1+g1 x / 2)and y∈(y1-g1 x / 2,y1+g1 x / 2)and z∈(z1-g1 x / 2,z1+g1 x / 2)

[0215] Number+1

[0216] Return table_deform(id,number)

[0217] Among them, id and number are the values ​​of table_cuboid_cell in the data analysis module, id is the number of the basic rectangular cell, and number is the number of points contained in the basic rectangular cell.

[0218] In graphics processing, this embodiment of the present invention scans and pre-processes complex spatial graphics in 3D, providing a foundation for subsequent digital processing. In the area division process, division rules can be set based on the actual object shape, and the area size can be adjusted to achieve accuracy. This system can also adapt to the recognition of various spatial graphics, with flexible parameter settings and machine learning capabilities. In polygon calculation, grid points are numbered and classified by region, allowing for the positional relationship of the smallest triangles to be determined at all times during statistics, and the total area of ​​each small triangle can also be calculated. This solution can quickly and accurately detect real-time changes in target objects, facilitating timely responses to these changes.

[0219] like Figure 3 As shown, an embodiment of the present invention provides a deformation recognition device 300, comprising:

[0220] An acquisition module 310 is configured to acquire at least two sets of three-dimensional point cloud data of a target object, wherein different sets of three-dimensional point cloud data correspond to different states of the target object;

[0221] The first processing module 320 is used to compress each set of 3D point cloud data to obtain compressed 3D point cloud data;

[0222] The second processing module 330 is used to convert and calculate the compressed three-dimensional point cloud data to obtain two-dimensional grid data;

[0223] The third processing module 340 is configured to divide the compressed three-dimensional point cloud data to obtain a plurality of divided first data units containing the three-dimensional point cloud data.

[0224] The fourth processing module 350 is configured to perform deformation recognition processing based on the relevant data of the first data unit and the two-dimensional grid data to determine a deformation result.

[0225] Optionally, the acquisition module 310 includes:

[0226] The acquisition submodule is used to scan the target object in different states multiple times using a 3D scanning device and pre-process the point cloud data obtained by the scans to obtain a set of three-dimensional point cloud data corresponding to the target object in each state.

[0227] Optionally, the first processing module 320 includes:

[0228] a first segmentation submodule, configured to segment the target three-dimensional point cloud data to obtain a plurality of second data units containing three-dimensional point cloud data; wherein the target three-dimensional point cloud data is any one set of three-dimensional point cloud data among the at least two sets of three-dimensional point cloud data;

[0229] A first calculation submodule, configured to calculate the curvature corresponding to the second data unit;

[0230] The compression submodule is used to compress the three-dimensional point cloud data in the second data unit according to the curvature corresponding to the second data unit to obtain compressed three-dimensional point cloud data.

[0231] Optionally, the first calculation submodule includes:

[0232] a first determining unit, configured to determine a target plane based on any three points contained in a target data unit, wherein the target data unit is any one of the plurality of second data units;

[0233] a first calculation unit, configured to calculate the distance between each point included in the target data unit and the target plane;

[0234] The second calculation unit is configured to calculate a standard deviation corresponding to the distance between each point and the target plane, and determine the standard deviation as the curvature.

[0235] Optionally, the compression submodule includes:

[0236] a first compression unit, configured to retain, for a second data unit having a curvature smaller than a first threshold, a point in the second data unit that is closest to a center of the second data unit;

[0237] a second compression unit, configured to retain, for a second data unit having a curvature greater than or equal to a first threshold, a plurality of points in the second data unit based on a curvature reduction algorithm;

[0238] The second determining unit is configured to determine the points retained in each second data unit as compressed three-dimensional point cloud data.

[0239] Optionally, the second data unit is a cuboid containing three-dimensional point cloud data, and the first threshold is determined based on the minimum value of the length, width and height of the cuboid.

[0240] Optionally, the second processing module 330 includes:

[0241] A first projection submodule is used to perform a first projection process on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data;

[0242] A second projection submodule is configured to obtain triangular mesh data having a connection relationship from the two-dimensional point cloud data based on a local greedy projection triangulation algorithm;

[0243] The determination submodule is configured to determine the number of triangular meshes and the total area of ​​the triangular meshes based on the triangular mesh data, and determine the number of triangular meshes and the total area of ​​the triangular meshes as the two-dimensional mesh data.

[0244] Optionally, the first projection submodule includes:

[0245] a third determining unit, configured to determine the center coordinates of the compressed three-dimensional point cloud data in a three-dimensional coordinate system;

[0246] a dividing unit, configured to divide the compressed three-dimensional point cloud data into a plurality of third data units containing three-dimensional point cloud data based on the center coordinates;

[0247] an establishing unit, configured to determine a projection surface based on each of the third data units, and establish a two-dimensional coordinate system based on the projection surface;

[0248] a fourth determining unit, configured to determine the three-dimensional projection coordinates of each point in the third data unit on the projection plane corresponding to the third data unit;

[0249] The fifth determining unit is used to determine the two-dimensional coordinates corresponding to the points in each third data unit based on the two-dimensional coordinate system and the three-dimensional projection coordinates, and determine the projected two-dimensional point cloud data based on the two-dimensional coordinates.

[0250] Optionally, the fourth processing module 350 includes:

[0251] A first processing submodule is configured to determine a degree of deformation according to the number of triangular meshes in the two-dimensional mesh data corresponding to the three-dimensional point cloud data;

[0252] a second processing subunit, configured to determine a deformation rate according to a total area of ​​a triangular mesh in the two-dimensional network data corresponding to the three-dimensional point cloud data;

[0253] a third processing submodule, configured to determine a deformation position according to the number of points contained in the first data unit corresponding to the three-dimensional point cloud data;

[0254] The fourth processing submodule is configured to determine a deformation result according to the deformation degree, the deformation rate, and the deformation position.

[0255] The device in the embodiment of the present invention can implement each process in the above-mentioned deformation recognition method and achieve the same technical effect. To avoid repetition, it will not be described here.

[0256] An embodiment of the present invention provides an electronic device, such as Figure 4 As shown, it includes a transceiver 410, a processor 400, a memory 420, and a program or instruction stored in the memory 420 and executable on the processor 400; when the processor 400 executes the program or instruction, the steps in the above-mentioned deformation recognition method are implemented and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0257] The transceiver 410 is configured to receive and send data under the control of the processor 400 .

[0258] Among them, Figure 4 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 400 and memory represented by memory 420. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are all well known in the art and, therefore, will not be described further herein. The bus interface provides an interface. The transceiver 410 may be a plurality of components, i.e., a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium. The processor 400 is responsible for managing the bus architecture and general processing, and the memory 420 may store data used by the processor 400 when performing operations.

[0259] A readable storage medium according to an embodiment of the present invention stores a program or instruction thereon. When the program or instruction is executed by a processor, the steps in the deformation recognition method described above are implemented and the same technical effect can be achieved. To avoid repetition, they will not be described here.

[0260] The processor is the processor described in the above embodiment. The readable storage medium includes a computer-readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0261] It should be further noted that the terminals described in this specification include but are not limited to smartphones, tablet computers, etc., and many functional components described are referred to as modules in order to more particularly emphasize the independence of their implementation methods.

[0262] In embodiments of the present invention, modules can be implemented in software so that they can be executed by various types of processors. For example, an identified executable code module can include one or more physical or logical blocks of computer instructions, for example, which can be constructed as objects, procedures, or functions. Nevertheless, the executable code of the identified module does not need to be physically located together, but can include different instructions stored in different locations, which, when logically combined together, constitute the module and achieve the specified purpose of the module.

[0263] In fact, executable code module can be a single instruction or many instructions, and can even be distributed on a plurality of different code segments, distributed in the middle of different programs, and distributed across a plurality of memory devices.Similarly, operating data can be identified in the module, and can be implemented and organized in the data structure of any appropriate type according to any appropriate form.Described operating data can be collected as a single data set, or can be distributed in different locations (including on different storage devices), and can only be present on a system or network as an electronic signal at least in part.

[0264] When a module can be implemented using software, given the current state of hardware technology, those skilled in the art can build corresponding hardware circuits to implement the corresponding functions of the module, regardless of cost. The hardware circuits may include conventional very large scale integration (VLSI) circuits or gate arrays, as well as existing semiconductors such as logic chips and transistors, or other discrete components. Modules may also be implemented using programmable hardware devices, such as field programmable gate arrays, programmable array logic, or programmable logic devices.

[0265] The above exemplary embodiments are described with reference to the accompanying drawings. Many different forms and embodiments are possible without departing from the spirit and teachings of the present invention. Therefore, the present invention should not be construed as limited to the exemplary embodiments set forth herein. Rather, these exemplary embodiments are provided so that this disclosure will be complete and perfect and will convey the scope of the invention to those skilled in the art. In the drawings, component sizes and relative sizes may be exaggerated for clarity. The terminology used herein is for purposes of describing specific exemplary embodiments only and is not intended to be limiting. As used herein, the singular forms "a," "an," and "the" are intended to include plural forms, unless the context clearly indicates otherwise. It will be further understood that the terms "comprising" and / or "including," when used in this specification, indicate the presence of stated features, integers, steps, operations, components, and / or elements, but do not preclude the presence or addition of one or more other features, integers, steps, operations, components, elements, and / or groups thereof. Unless otherwise indicated, when stated, a range of values ​​includes the upper and lower limits of that range and any subranges therebetween.

[0266] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A deformation recognition method, characterized in that: include: Acquire at least two sets of three-dimensional point cloud data of a target object; wherein different sets of three-dimensional point cloud data correspond to different states of the target object; Performing compression processing on each set of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data; Perform conversion and calculation on the compressed three-dimensional point cloud data to obtain two-dimensional grid data; performing segmentation processing on the compressed three-dimensional point cloud data to obtain a plurality of segmented first data units containing the three-dimensional point cloud data; Performing deformation recognition processing based on the three-dimensional point cloud data of the first data unit and the two-dimensional grid data to determine a deformation result; The step of compressing each set of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data includes: Dividing the target three-dimensional point cloud data to obtain a plurality of second data units containing three-dimensional point cloud data; wherein the target three-dimensional point cloud data is any one set of three-dimensional point cloud data among the at least two sets of three-dimensional point cloud data; calculating a curvature corresponding to the second data unit; The three-dimensional point cloud data in the second data unit is compressed according to the curvature corresponding to the second data unit to obtain compressed three-dimensional point cloud data.

2. The deformation recognition method according to claim 1, characterized in that: The step of obtaining at least two sets of three-dimensional point cloud data of the target object includes: For the target object in different states, the target object is scanned multiple times by a 3D scanning device and the point cloud data obtained by the scanning is preprocessed to obtain a set of three-dimensional point cloud data corresponding to the target object in each state.

3. The deformation recognition method according to claim 1, characterized in that: The calculating the curvature corresponding to the second data unit includes: Determining a target plane based on any three points contained in a target data unit; wherein the target data unit is any one of the plurality of second data units; Calculating the distance between each point contained in the target data unit and the target plane; A standard deviation corresponding to the distance between each point and the target plane is calculated, and the standard deviation is determined as the curvature.

4. The deformation recognition method according to claim 1, characterized in that: The compressing the three-dimensional point cloud data in the second data unit according to the curvature corresponding to the second data unit to obtain the compressed three-dimensional point cloud data includes: For a second data unit whose curvature is smaller than a first threshold, retaining a point in the second data unit that is closest to the center of the second data unit; For a second data unit having a curvature greater than or equal to a first threshold, retaining a plurality of points in the second data unit based on a curvature reduction algorithm; The points retained in each second data unit are determined as compressed three-dimensional point cloud data.

5. The deformation recognition method according to claim 4, characterized in that: The second data unit is a cuboid containing three-dimensional point cloud data, and the first threshold is determined based on the minimum value of the length, width, and height of the cuboid.

6. The deformation recognition method according to claim 1, characterized in that: The step of converting and calculating the compressed three-dimensional point cloud data to obtain two-dimensional grid data includes: Performing a first projection process on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data; Based on a local greedy projection triangulation algorithm, triangular mesh data with a connection relationship is obtained from the two-dimensional point cloud data; Based on the triangular mesh data, the number of triangular meshes and the total area of ​​the triangular meshes are determined, and the two-dimensional mesh data is determined according to the number of triangular meshes and the total area of ​​the triangular meshes.

7. The deformation recognition method according to claim 6, characterized in that: The performing a first projection process on the compressed three-dimensional point cloud data to obtain projected two-dimensional point cloud data includes: Determining the center coordinates of the compressed three-dimensional point cloud data in a three-dimensional coordinate system; Based on the center coordinates, dividing the compressed three-dimensional point cloud data into a plurality of third data units containing three-dimensional point cloud data; determining a projection surface based on each of the third data units, and establishing a two-dimensional coordinate system based on the projection surface; Determine the three-dimensional projection coordinates of each point in the third data unit on the projection plane corresponding to the third data unit; Based on the two-dimensional coordinate system and the three-dimensional projection coordinates, the two-dimensional coordinates corresponding to the points in each third data unit are determined, and the projected two-dimensional point cloud data is determined based on the two-dimensional coordinates.

8. The deformation recognition method according to claim 6, characterized in that: The performing deformation recognition processing based on the three-dimensional point cloud data of the first data unit and the two-dimensional grid data to determine the deformation result includes: determining a degree of deformation according to the number of triangular meshes in the two-dimensional mesh data corresponding to the three-dimensional point cloud data; determining a deformation rate according to a total area of ​​a triangular mesh in the two-dimensional network data corresponding to the three-dimensional point cloud data; determining a deformation position according to the number of points included in the first data unit corresponding to the three-dimensional point cloud data; A deformation result is determined according to the deformation degree, the deformation rate, and the deformation position.

9. A deformation recognition device, characterized in that: include: An acquisition module, configured to acquire at least two sets of three-dimensional point cloud data of a target object; wherein different sets of three-dimensional point cloud data correspond to different states of the target object; The first processing module is used to compress each set of three-dimensional point cloud data to obtain compressed three-dimensional point cloud data; The second processing module is used to convert and calculate the compressed three-dimensional point cloud data to obtain two-dimensional grid data; a third processing module, configured to divide the compressed three-dimensional point cloud data to obtain a plurality of divided first data units containing the three-dimensional point cloud data; a fourth processing module, configured to perform deformation recognition processing based on the three-dimensional point cloud data of the first data unit and the two-dimensional grid data to determine a deformation result; Wherein, the first processing module includes: a first segmentation submodule, configured to segment the target three-dimensional point cloud data to obtain a plurality of second data units containing three-dimensional point cloud data; wherein the target three-dimensional point cloud data is any one set of three-dimensional point cloud data among the at least two sets of three-dimensional point cloud data; A first calculation submodule, configured to calculate the curvature corresponding to the second data unit; The compression submodule is used to compress the three-dimensional point cloud data in the second data unit according to the curvature corresponding to the second data unit to obtain compressed three-dimensional point cloud data.

10. An electronic device comprising: A transceiver, a processor, a memory, and a program or instruction stored in the memory and executable on the processor; characterized in that when the processor executes the program or instruction, the steps in the deformation recognition method as described in any one of claims 1 to 8 are implemented.

11. A readable storage medium having a program or instruction stored thereon, characterized in that: When the program or instruction is executed by a processor, the steps of the deformation recognition method according to any one of claims 1 to 8 are implemented.

Images